Re: bending and radius of curvature

Posted by Gabriel Landini on Dec 21, 2010; 11:42am
URL: http://imagej.273.s1.nabble.com/Re-bending-and-radius-of-curvature-tp3686117p3686121.html

On Tuesday 21 December 2010 11:49:10 you wrote:
> For closed shapes, I rather would associate "Fit Circle" to be similar to
> "Fit Ellipse", i.e. same area and centroid. This is currently not the case:
> a longish, symmetrical particle outline could result in a very large radius
> (see macro below). Also if I remove the "Fit Spline" in the macro, I don't
> understand the meaning of the resulting circle.

I guess that it is assuming a cloud of points (in this case the points forming
the ellipse) as defining an arc of the circle (instead of the whole circle).
Not sure how the algorithm works but maybe it looks at minimising residuals.
If so, the arc might have smaller residuals than a circle which intuitively
looks similarly sized to the cloud?


Cheers

Gabriel